Keyframe-based monocular SLAM: design, survey, and future directions

Younes, Georges; Asmar, Daniel; Shammas, Elie; Zelek, John

doi:10.1016/j.robot.2017.09.010

Cited by 149 publications

(77 citation statements)

References 104 publications

(132 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This involves a generic Structurefrom-Motion ( [28]) based on local bundle adjustment, which provides keyframes and matched Harris points that are used latter. We also reduce the accumulated drift of the trajectory thanks to detection and closure of loops inspired by [45,44]: detection using a vocabulary tree of the matched points, closure using global bundle adjustment initialized by pose graph optimizations.…”

Section: Sparse Input Point Cloud From Imagesmentioning

confidence: 99%

Surface reconstruction from a sparse point cloud by enforcing visibility consistency and topology constraints

Lhuillier

2018

Computer Vision and Image Understanding

View full text Add to dashboard Cite

Highlights• Surface reconstruction under-uses topology constraints in Computer Vision• Start from a previous method enforcing manifoldness on a sparse point cloud• Simultaneously enforce visibility consistency and low genus for the first time• Improve the removal of surface singularities and the escape from local extrema• Experiment on long video sequences taken by a helmetheld omnidirectional camera AbstractThere are reasons to reconstruct a surface from a sparse cloud of 3D points estimated from an image sequence: to avoid computationally expensive dense stereo, e.g. for applications that do not need high level of details and have limited resources, or to initialize dense stereo in other cases. It is also interesting to enforce topology constraints (like manifoldness) for both surface regularization and applications. In this article, we improve by several ways a previous method that enforces the manifold constraint given a sparse point cloud. We enforce lowered genus, i.e. simplified topology, as a further regularization constraint for maximizing the visibility consistency encoded in a 3D Delaunay triangulation of the points. We also provide more efficient escapes from local extrema, an acceleration of the manifold test and more efficient removals of surface singularities. We experiment on a sparse point cloud reconstructed from videos, that are taken by a helmet-held omnidirectional multi-camera moving in an university campus.

show abstract

Section: Sparse Input Point Cloud From Imagesmentioning

confidence: 99%

Surface reconstruction from a sparse point cloud by enforcing visibility consistency and topology constraints

Lhuillier

2018

Computer Vision and Image Understanding

View full text Add to dashboard Cite

show abstract

“…LSD-SLAM [5] is another state-of-the-art Monocular SLAM system which works with image intensities and creates a semi-dense map of the environment. A survey of various Monocular SLAM methods can be found in [26].…”

Section: Background 21 Monocular Slammentioning

confidence: 99%

Learning to Prevent Monocular SLAM Failure using Reinforcement Learning

Prasad¹,

Yadav

Saurabh

et al. 2018

Proceedings of the 11th Indian Conference on Computer Vision, Graphics and Image Processing

View full text Add to dashboard Cite

Monocular SLAM refers to using a single camera to estimate robot ego motion while building a map of the environment. While Monocular SLAM is a well studied problem, automating Monocular SLAM by integrating it with trajectory planning frameworks is particularly challenging. This paper presents a novel formulation based on Reinforcement Learning (RL) that generates fail safe trajectories wherein the SLAM generated outputs do not deviate largely from their true values. Quintessentially, the RL framework successfully learns the otherwise complex relation between perceptual inputs and motor actions and uses this knowledge to generate trajectories that do not cause failure of SLAM. We show systematically in simulations how the quality of the SLAM dramatically improves when trajectories are computed using RL. Our method scales effectively across Monocular SLAM frameworks in both simulation and in real world experiments with a mobile robot.

show abstract

“…Environment dynamics are essential as most algorithms for indoor do not scale well for outdoor applications. Regarding theoretical frameworks, SLAM can be categorised into two main paradigms: filtering and optimization based approach [26]. Some of the filtering methods are: Extended Kalman Filter (EKF) and Rao-Blackwellized particle filters (FastSLAM) [27].…”

Section: B Representations For Robot Localization and Mapping: Back-endmentioning

confidence: 99%

Fire detection of Unmanned Aerial Vehicle in a Mixed Reality-based System

Esfahlani

Cirstea

Sanaei

et al. 2018

IECON 2018 - 44th Annual Conference of the IEEE Industrial Electronics Society

View full text Add to dashboard Cite

This paper presents a system that combines robotic operating system (ROS) and computer vision techniques for fire detection in a mixed reality environment. We have collected video streams from a mini camera on an Unmanned Aerial Vehicle (UAV), where the navigation data relied on state-of-theart Simultaneous Localization and Mapping (SLAM) system. The data collected onboard are communicated to the ground station and processed through the robotic operating system. A robust and efficient re-localisation SLAM was performed to recover from tracking failure and frame lost in the received data. The fire detection algorithm was developed based on the colour, movement attributes, temporal variation of fire intensity and its accumulation around a point. A mixed reality environment was used to visualise and test the proposed system. The observation and data analysis confirmed that the UAV could successfully detect fire and flame, fly towards and hover around it.

show abstract

Keyframe-based monocular SLAM: design, survey, and future directions

Cited by 149 publications

References 104 publications

Surface reconstruction from a sparse point cloud by enforcing visibility consistency and topology constraints

Surface reconstruction from a sparse point cloud by enforcing visibility consistency and topology constraints

Learning to Prevent Monocular SLAM Failure using Reinforcement Learning

Fire detection of Unmanned Aerial Vehicle in a Mixed Reality-based System

Contact Info

Product

Resources

About